Geocloud is a simple command line tool to combine LiDAR data and trajectory data to create a fully georeferenced 3D pointcloud. It takes a PCAP file collected with a Velodyne LiDAR scanner synchronised with an OxTS inertial navigation system (INS), a processed NCOM file from the INS, along with a few configuration files to create a LAS file that can be viewed in many 3rd party LiDAR software packages.
Geocloud requires data from synchronised Velodyne and OxTS systems.
GeoCloud is still currently in beta. Email us here to request access.
Processing INS data
The INS raw data (*.rd) file must be processed with OxTS’s post-processing application NAVsolve to create an NCOM file. Some of the processing configuration files are required by GeoCloud, so the option to clean up directory after processing in the Advanced Settings should be left unchecked.
The RD file also needs to be processed with local coordinates enabled. If local coordinates were not configured before the data collection, they can be enabled for post-processing by following these steps:
Load the RD file with NAVsolve
Loading an RD file in NAVsolve automatically extracts the config files stored in the RD file. Select the RD file on the Source page and set the working directory. Moving to the Preview page will create a folder called ExtractedConfiguration in the working directory.
Create an LRF file
An LRF file is a configuration file that defines the local coordinate system. It is a simple text file of four lines with one number on each line:
WGS84 altitude (m)
heading of x-axis (degrees)
Save the file with the same filename as the RD file you wish to apply it to. Change the file extension from .txt to .lrf.
Store the LRF file in the config folder
Navigate to the ExtractedConfiguration folder and copy the LRF file to the folder. Make sure the filename is the same as the other config files there.
Process the file
Continue through NAVsolve and process the file. The resulting NCOM file will contain local coordinate data.
GeoCloud is used through a command line interface. When running
geocloud.exe, it looks for a default set of required files in the directory. Alternatively, you can use a number of command line arguments to specify the name and location of the required files, and configure some advanced options. The table below describes the required files, default names, and advanced options.
||mobile.pcap||LiDAR Data filename|
||mobile.ncom||NCOM data filename|
||mobile.lip||Position of LIDAR in INS frame (in metres)||x (m)
||mobile.lir||Attitude of LIDAR in INS frame (in °)||h (°)
||mobile.vat||Attitude of INS in Vehicle frame (in °)||h (°)
||+||Enable LAS output|
||-||Enable PCD output|
||-||Enable LAS compression (LAZ)|
||8686080||Number of points in cloud before writing PCD file to disk|
||0||Until processed (pcd_pcloud_size*skip_first_n_pclouds), don’t write pclouds to disk|
||256||After processing (pcd_pcloud_size*drop_plcouds_after) points, stop writing pclouds|
||1||GPS nano time before which no LCOM packets are to be processed|
||calib.csv||(Calibration) Output file for calibration points (see XXX)|
||255||(Calibration) Any points with reflectivity > ref_threshold will be included in cal_file (For VLP-16, set this to 100 to include all retro-reflective returns)|
The VAT file can be found in the ExtractedConfiguration folder created when processing the RD file.
The LIP and LIR files define the boresight alignment of the INS-LiDAR system. They can be created using a simple text editor such as Notepad.
geocloud.exe from the command line will produce a fully georeferenced 3D pointcloud in the LAS format, which can then be dropped into many 3rd party applications for further processing and analysis. There are also a number of free, simple LAS viewers available such as CloudCompare.
The LIP file should contain 3 lines with a number on each describing the x, y, and z offsets from the INS to LiDAR measurements points in the INS frame. The measurements should be made as accurate as possible, within 1 cm.
The LIR file can be created by performing a boresight alignment calibration run (article coming soon). Otherwise it should contain 3 lines with a number on each describing the roll, pitch, and yaw offsets from the INS to the LiDAR. If performing a calibration, the initial measurements do not need to be accurate as the alignment converges regardless of the initial inputs.